基于对数累积量的重尾分布脉冲干扰参数估计及性能分析

发布时间：2018-07-03 11:42

本文选题：对数累积量 + 重尾Rayleigh分布　；参考：《中国科学:信息科学》2017年02期

【摘要】：在无线通信网络中存在用对称α稳定分布来建模的脉冲重尾干扰.而在信号检测、信道译码、无线网络中断概率及误码率分析等应用场景,需要预先知道干扰的概率密度函数.本文利用重尾干扰复信号包络的对数累积量,给出了特征指数和分散系数的估计算法,并具体推导出了参数估计变量的概率分布,该分布可用于定量分析估计的可靠性.除此之外,在实际系统中,接收端不仅有复对称α稳定分布描述的重尾脉冲干扰,还包括与之相互独立的复Gauss噪声,称之为双变量混合噪声.本文提出了用单变量的复对称α稳定分布模型来近似双变量混合噪声的方法.通过仿真和数值计算,验证了这种近似是合理的.再者,在此基础上,本文给出了混合噪声参数与几何功率信噪比之间的关系.因此,在合理的近似下,对数累积量的估计算法及性能分析在双变量混合噪声下仍然有效.
[Abstract]:In wireless communication networks, there is a pulse heavy-tailed interference modeled by symmetric 伪 stable distribution. In signal detection channel decoding wireless network interrupt probability and bit error rate analysis we need to know the probability density function of interference in advance. Based on the logarithmic cumulant of the envelope of heavy-tailed interference complex signals, the estimation algorithms of characteristic index and dispersion coefficient are given, and the probability distribution of parameter estimation variables is derived in detail, which can be used to quantitatively analyze the reliability of estimation. In addition, in the actual system, the receiver not only has the heavy-tailed pulse interference described by the complex symmetric 伪 stable distribution, but also includes the complex Gauss noise which is independent of the complex Gaussian noise, which is called bivariate mixed noise. In this paper, a method of approximating bivariate mixed noise by using the complex symmetric 伪 stable distribution model of single variable is presented. The simulation and numerical calculation show that this approximation is reasonable. Furthermore, the relationship between mixed noise parameters and geometric power signal-to-noise ratio is given. Therefore, under reasonable approximation, the estimation algorithm and performance analysis of logarithmic cumulants are still valid under bivariate mixed noise.
【作者单位】：浙江警察学院计算机与信息技术系;南方科技大学电子与电气工程系;浙江工业大学信息工程学院;浙江省通信网技术应用研究重点实验室;
【基金】：国家自然科学基金浙江两化融合联合基金(批准号:U1509219) 国家自然科学基金(批准号:61471322,61402416)资助项目
【分类号】：O211;TN911.4

【相似文献】